Discussion

The current IMWG definition of MM-related RI requires a serum creatinine of 2?g/dl or higher that is unexplainable by any other etiology.²⁵ However, this cutoff may fall short in identifying all NDMM patients with RI since the serum creatinine can be influenced by factors such as muscle mass. Conversely, a cutoff of CrCl <60?ml/min may inappropriately consider older NDMM patients with age-related decline in their CrCl as having MM-related RI. In this study, we used a CrCl cutoff of <40?ml/min as it has appeared to be optimal in identifying NDMM patients with RI when compared to the previously mentioned cutoffs.²⁶

By using our CrCl cutoff of 40?ml/ml, RI at diagnosis was present in almost 20% of the patients. Of these patients, more than half of them (54%) had reversal of their RI upon institution of anti-myeloma induction therapy. This study demonstrated an improvement in survival among NDMM patients with RI as a result of novel agent therapy as seen in Figure 2. Furthermore this study suggests that though RI in NDMM patients is associated with a worse OS as well as higher rates of early mortality, it is not an independent predictor of worse OS as seen in the multivariable model (Table 2). However, this study demonstrates that even if NDMM patients with RI experience a resolution of their RI upon receiving myeloma directed therapy (group 2), they do not have equivalent survival outcomes to those NDMM patients without RI (group 1) as seen in Figure 3a. Furthermore, even after accounting for early mortality by performing a landmark analysis at 6 months, patients in group 2 still do not have equivalent survival outcomes as those patients in group 1 (Figure 2). However, reversal of RI is still important to achieve in NDMM patients. This is because patients in group 3 who have RI at diagnosis but never recover their renal function have a worse median OS to patients in group 2 (33 vs 56 months, P=0.006). This survival benefit of reversal of RI seen in group 2 is observed likely due to their significantly lower rate of early mortality in comparison to group 3 (8 vs 17%; P=0.004). When a landmark analysis was performed at 6 months, the median OS for patients in group 2 compared to group 3 was 67 vs 51 months (P=0.175).

Our study also evaluated the renal function of all the patients at their last follow-up. We observed that even though patients in group 2 had a reversal of their RI, they were more likely to eventually re-worsen their renal function and experience RI again compared to the patients in group 1 (34 vs 9%; P<0.001); this suggests that NDMM patients with RI at diagnosis likely have a higher propensity for RI during their disease course compared to their NDMM counterparts who do not have RI at diagnosis.

Several studies have confirmed the beneficial effect of novel agent induction therapy in NDMM patients with RI in comparison to conventional chemotherapy.^{27, 28} This is likely because novel agents have been associated with improved depth of paraprotein response in myeloma and this likely translates to higher rates of improvement in renal function.^{29, 30, 31, 32} RI at diagnoses has not been found to weaken the responses provided by novel agent induction therapy.³³ Furthermore, unlike most conventional chemotherapeutic agents, novel agents such as bortezomib and thalidomide have safe pharmacokinetic and pharmacodynamic properties in the setting of severe RI;³⁴ lenalidomide, even though it is mainly renally excreted, can also be safely used if appropriate dose modifications are implemented and close evaluation for toxicities are undertaken in patients with mild to moderate RI.³⁵ Novel agents such as bortezomib have also been reported to have protective effects in the renal tubular cells³⁶ and inhibitory effects on the pro-inflammatory and fibrotic pathways within the renal microenvironment³⁷ in addition to its previously described anti-myeloma activity.

However, despite the survival gains experienced with the emergence of novel agents in the upfront management of MM patients with RI at diagnosis, there is still room for improvement. Acute renal failure has been a major cause of early mortality in previous observational studies involving NDMM patients.^{9, 38} Even in this study, we observed a 16% early mortality rate in NDMM patients with CrCl<40?ml/min treated with a novel agent induction regimen; this signifies the need for earlier and more effective interventions in this subgroup of patients.

Our study also indicated that certain patient and disease-related characteristics predicted for the presence of RI at diagnosis in NDMM patients. Age >70 years was the only patient characteristic associated with RI at diagnosis. However, higher disease burden as suggested by an international staging system 3, high-risk cytogenetics by FISH, light-chain-secreting-only disease, higher bone marrow plasma cell percentage and higher LDH were all determinants of plasma cell biology that were associated with RI at diagnosis. In contrast, only NDMM patients who did not have light-chain-secreting-only disease were most likely to recover their renal function. In this study, older age, high-risk FISH, elevated LDH, high plasma cell labeling index and lack of novel agent use during induction were independent predictors of worse OS. This suggests that RI at diagnosis may not be an adverse prognostic marker in NDMM patients. In a study by Eleftherakis-Papapiakovou et al.,¹⁹ RI was also not found to be independently associated with inferior survival likely as a result of novel agent use during induction therapy.

There are several limitations to this study. First, the etiology of RI in each patient was not included in the analyses. Given the advanced age of the average NDMM patient, it is possible that their RI can also be associated with a decline in renal function as a result of other medical comorbidities such as diabetes, hypertension, vascular disease, drug-induced issues and so on. unrelated to the MM. Furthermore, a renal biopsy is not necessarily required or performed in every NDMM patient with RI¹¹ and we do not have that information on all of our patients making it difficult to know the true etiology of RI in our patient population. Second, we are unable to determine with certainty whether certain novel agents used during induction such as bortezomib is superior to immunomodulators like thalidomide or lenalidomide in this subpopulation with RI due to the potential bias in patient selection for certain therapies and the high likelihood of most patients receiving bortezomib in the salvage setting. Third, the retrospective nature of this study prevents us from truly understanding the etiology of early mortality in these patients, that is, disease-related morbidity vs therapy-related toxicity leading to early mortality. Also in patients with acute renal impairment, the traditional equations to calculate glomerular filtration rate such as the MDRD formula do not provide accurate assessments of their true CrCl.³⁹ Nevertheless, the MDRD equation has been adopted by the IMWG in monitoring renal function and response in patients with newly diagnosed MM. ¹¹

Data from randomized control trials comparing different induction therapy regimens in NDMM patients with RI are sparse. Nevertheless, our data confirm the improvement in renal response that novel agents have had on NDMM patients with RI. They have also decreased early mortality in these patients as well as improved OS. However, reversal of RI alone does not appear to elevate the expected OS of such NDMM patients with RI at diagnosis to that of patients without RI at diagnosis. The findings from this study imply the need for instituting early treatment strategies in order to prevent patients from developing RI. For example, redefining the existing CRAB criteria required to diagnose MM by including criteria such as free light chain ratio >100 (ref. 40) or a bone marrow plasma cell percentage of 60 or higher⁴¹ may identify NDMM patients requiring therapy prior to them developing RI. In addition, for those NDMM patients already with RI, further work is required to determine the optimal management so as to continue to reduce associated morbidity and mortality.

...